PTQ Q4 2023 Issue

the most important parameter for reducing losses. Denser catalysts may better retain 0-40 micron particles. Attrition measured with a jet cup is the number of fines produced when subjected to a high-velocity gas jet. The data shown in Figure 4 divide the attrition regimes into two domains. At lower velocities, the abrasion regime is the only type that occurs. This is the rubbing of particles or the mild abrasion that occurs with contact to solid surfaces. Higher velocities will shatter the particles, producing larger fragments. In the early days of catalytic cracking, units were designed with attrition nozzles that could pro- vide fines to aid in catalyst circulation because early cyclone systems did not have the efficiency of modern cyclones. The loss of fines in units with long standpipes (more than 60 ft long and some more than 100 ft) made the smooth operation of the standpipes difficult. Steam introduced at sonic velocity could increase the 0-40 particles from 0 to 25% in a day. Velocities through air and steam distributors should be less than 300 ft/sec to limit losses with typical catalysts used today. Feed nozzles should also have a maximum velocity of about 275 ft/sec to prevent increased losses. The degree of atomisation is important in the design, and some nozzles can attrit cata- lyst at much lower velocities. Most of the attrition that occurs in the FCC unit is abra- sive, and the size of the components making up the catalyst is significant. The zeolites and clay particles are 2 microns or less and usually comprise at least 50% of the catalyst. These can chip off the catalyst particle by the rubbing of catalyst particles and/or low-velocity impacts. Fracturing occurs when the velocities are high, resulting in two peaks in the fine catalyst particle size distribution. Missing or mis-sized restriction orifices or bypasses left open can be a source of high-velocity jets. Torch oil noz- zles have steam to assist in atomisation and to keep them from coking when they are not in use. If the bypass is left open, the velocity can be sonic through the nozzles. This is a prime source of high losses after a start-up. Wet steam entering a standpipe can attrit catalyst and disturb the catalyst flow. Emergency steam put in the riser will have high velocities. These nozzles are sized to give a superfi - cial velocity in the riser capable of circulating the catalyst when the feed is removed. A refiner found that this steam improved the yields, but catalyst losses increased 2-3 t/d. The tip velocity was about 500 ft/sec. In resid operations, the use of water injection into the regenerator bed was used, as well as steam coils for heat removal. Adding water to the bed will cause increased cat- alyst attrition and deactivate the catalyst. Steam coils are prone to leaking and have been replaced with dense-phase catalyst coolers. Catalyst attrition can occur when loading or unloading the catalyst. The lines should be designed for minimum pressure drop, with few bends and be as short as possi- ble. Erosion of the catalyst lines is common on FCC units, as attested by the scabs placed on the loading lines where severe erosion or holes developed. An equation for ero- sion in transfer lines is: K = (particle density x pipe dia. X gas velocity 3 )/radius pipe bend

Velocity is the main variable and should be limited to 40-50 ft/sec. A truck driver arriving in the late afternoon with a date will likely exceed these velocities. Severe attri- tion of the fresh catalyst can occur, and the catalyst will be lost soon after it is added to the unit. The losses from FCC units are very low when calculated from the efficiency of the basic separators in the reactor and regenerator. Typical efficiencies of two-stage cyclone systems are:

Two-stage cyclone efficiencies • Poor

<99.995 99.996 99.997

• Marginal

• Good

• Exceptional 99.998 These efficiency numbers are so high due to the impact of high loading in which the small particles are pinned against the cyclone walls and recovered. The losses from catalytic cracking units have been char- acterised as:

Loss rate Lb/B

Rating

0.04-0.05

Tight

0.07 0.10 0.12 0.15 0.20

Excellent

Good

Reasonable Borderline

Poor

Conclusions The losses can be identified quickly if the data from the FCC unit are taken and converted into useful information. Software should be developed that shows the losses from each side of the unit and the efficiency of the collection sys - tem. Changes in these numbers and variations in the par- ticle size distributions of the equilibrium catalyst and fines leaving the unit will pinpoint the area of concern. Possible remedies can be tried, and plans made for future actions. References 1 Fletcher R, Evans M, Optimizing & troubleshooting the FCC regenera- tor for reduced emissions , NPRA Annual Meeting, AM-19-173 2 Wilson J, Fluid Catalytic Cracking Technology and Operation, Pennwell Books. 3 Niccum P K, 20 Questions: Identify probable causes for high FCC Catalyst Loss, Hydrocarbon Processing , Sept. 2010, 29-38. 4 Zenz F A, Othmer D, Fluidization and Fluid Particle Systems , Reinhold Chemical Engineering Series, 1960. 5 Fletcher R, Stepwise method determines source of FCC catalyst loss, Oil and Gas Journal , 28 Aug 1995. 6 Rowe P N, Zenz F A, The Chemical Engineering Consultant: an industrial detective-Part II, Institute of Chemical Engineers, 16 Belgrave Square, London S.W.I. 7 Maholland M, McAuley D, Failure mechanism for FCC cyclones , NPRA Annual Mtg, AM-10-110. 8 Wojtowicz M, Mitigating catalyst losses , AFPM Q&A Conference, 6-8 Oct 2014. Warren Letzsch P.E. is a refinery consultant with expertise in FCC catal - ysis and process technology and more than 40 years’ experience in the refining and petrochemical industry.

PTQ Q4 2023